Tackling Information Infrastructure Complexity

IT information infrastructures are incredibly complex, and expected increased application demands and data growth will only exacerbate that complexity. Although no Alexander the Great exists to cut the complexity Gordian Knot completely, IT organizations are investigating how to take a software-driven and data-oriented approach to at least make that complexity more manageable. Sanbolic's Melio software is an example of the options available for that purpose.

IT Challenged By Complexity

Information infrastructures grow incrementally, in fits and starts, to meet individual application needs. Hardware resources--servers, storage and networking requirements--tend to get added in at the end of the planning process. This results in application silos with their own dedicated, isolated resources. Now, to be fair, IT has long attempted to break down theses silos to some extent with shared storage systems (notably, SANs), which allow many applications to share the same physical array, and with server virtualization, which allows multiple applications to share the same physical system and/or CPU.

However, that is not enough. The increased mobility required to migrate virtual machines to a different physical server and storage to different tiers improves efficiency, but is only one of the new challenges. Another is the heterogeneous nature of IT resources--such as servers, storage and operating systems--which complicates the shared management of applications. On top of this is to the need to scale to meet increasing demands, especially with tight IT budgets.

Complexity results in two key problems: unacceptably high costs that are only going to grow worse as demands scale and increased pressure and threat to maintaining mandated service levels (such as high availability). This latter point is not only nice to have, but essential for delivering the continuous availability needed for generating revenues and ensuring customer retention.

[Enterprises store a lot of data that has little business value. David Hill looks at the problem in "The Critical Need For Data Disposal."]

Although it is by no means there yet, a consensus seems to be developing on how to ameliorate this IT complexity. Everyone is talking about a “software-defined” something or other (such as data center, storage, and networking), so software-defined everything seems to be the hot IT topic of note.

But that is not the only approach. Many IT vendors’ and customers’ focus has turned to data. An application may create data originally. Other applications, such as backup/recovery, are responsible for data protection, and other software tools are used for deriving further value from a secondary use of the data (which is what big data is all about). Focusing on applications seems good at the beginning, but cannot effectively attack complexity alone. Some have suggested the use of data-driven software intelligence to attack complexity as the best or only other alternative. For these people, software provides the intelligence to take the needed actions, and a data focus provides the target on which to take those actions.

That leads us to Sanbolic as an illustration of a software company that is attacking the heart of the IT complexity problem.

How Sanbolic Works

Founded in 2000, Sanbolic has more than 800 customers. But it has been only recently that the software-defined technologies and data focus that are in Sanbolic’s wheelhouse have attracted so much attention. Sanbolic is a software-only company, and Melio (currently in release 5) is its mature flagship data management software product. Nominally, Melio has at its core a clustered file system. Having a global namespace from such a file system is a solid asset, but Melio’s key value is that it touches and provides services for all servers and storage. That means it provides an overarching management control point for key resources in the information infrastructure. This helps reduce or even eliminate siloed servers and storage, and means that actions can be taken to manage everything as a whole (a prerequisite to the cloud).

Melio 5 is also host-based, a key attribute in the term software-defined. That means the software location is decoupled from the location of the physical hardware that it manages. But data management is really about the non-data path control and use of data throughout its lifecycle.

Melio's goal is to improve scalability and availability of applications. For example, as a subset of its overall data management capabilities, Melio 5 can provide software-defined storage. Software-defined storage is about using host-based software to manage and control multiple, heterogeneous storage arrays for provisioning, virtualization and other services.

But how does such data management software apply to applications? Although applications should no longer be the tail that wags the IT information infrastructure dog, they still maintain a central and absolutely critical function. In fact, the data-defined software lens that Melio employs better meets applications’ needs for both scale-out and availability. Plus, Melio’s ability to consolidate, cluster and manage physical, virtual and cloud workloads enables more agile, efficient and scalable application deployments.

When applications are siloed, resource utilization is often poor (that is, inefficient), barriers to moving to a more efficient environment exist (even in a server virtualized environment), and adding resources to meet unexpected demand is not easy, so it’s definitely not scalable.

Melio’s ability to add computing nodes non-disruptively online in a cluster is key to enabling these critical services. Melio 5 also provides volume management and a number of other capabilities to provide high availability to applications and data. In addition to standard approaches, such as RAID, Melio includes some clever confidential approaches that go into the recovery of data.

NEXT: Melio In The Cloud